Table 5

Logit model

DEGEIGENBETCLOSPCAN-SCORE
CENTRALITY0.371**0.429**0.190*−0.05450.216**0.424**
(0.163)(0.182)(0.114)(0.0656)(0.0933)(0.193)
SIZE−0.435***−0.405***−0.355**−0.357***−0.410***−0.380***
(0.140)(0.137)(0.139)(0.137)(0.140)(0.139)
LEV1.382*1.2291.404*1.384*1.3121.179
(0.833)(0.824)(0.833)(0.828)(0.834)(0.833)
ROA−2.591−2.404−2.810−2.601−2.875−2.607
(2.348)(2.309)(2.327)(2.318)(2.339)(2.330)
EXP0.2010.09290.1980.2330.1410.116
(0.569)(0.574)(0.549)(0.561)(0.568)(0.565)
RECEI−2.941**−2.820**−2.742**−2.795**−2.802**−2.822**
(1.228)(1.259)(1.268)(1.255)(1.261)(1.272)
LOSS−0.341−0.334−0.363−0.359−0.352−0.326
(0.321)(0.327)(0.318)(0.318)(0.322)(0.321)
LOSSt−1−0.142−0.142−0.142−0.129−0.174−0.175
(0.296)(0.293)(0.287)(0.290)(0.296)(0.295)
CLEANt−13.055***3.002***3.061***3.061***3.053***3.034***
(0.221)(0.221)(0.224)(0.225)(0.223)(0.223)
AC_SIZE0.04170.04250.08990.06220.06810.0769
(0.0615)(0.0616)(0.0621)(0.0611)(0.0614)(0.0615)
AC_MEET−0.00332−0.0004980.006220.003160.003360.00283
(0.0536)(0.0527)(0.0508)(0.0517)(0.0520)(0.0517)
B_MEET−0.00543−0.00778−0.0144−0.0148−0.00893−0.0109
(0.0350)(0.0353)(0.0340)(0.0345)(0.0341)(0.0344)
B_INDEP1.988***1.729***1.758***1.737***1.835***1.858***
(0.516)(0.489)(0.477)(0.484)(0.489)(0.484)
BIG41.478***1.466***1.482***1.484***1.485***1.437***
(0.381)(0.366)(0.376)(0.372)(0.378)(0.372)
AFt−1−0.0928−0.0807−0.0845−0.0772−0.0905−0.0743
(0.131)(0.125)(0.126)(0.124)(0.129)(0.127)
RLAW0.3150.2630.3610.3420.3120.311
(0.234)(0.236)(0.228)(0.230)(0.230)(0.229)
REST_LINKS−0.0104−0.0183−0.00167−0.00748−0.00519−0.00673
(0.0747)(0.0734)(0.0751)(0.0737)(0.0745)(0.0748)
LINKS_TOTAL−0.104−0.0587−0.0952−0.0386−0.119−0.0994
(0.0742)(0.0641)(0.0772)(0.0681)(0.0766)(0.0706)
Intercept2.5992.3911.1641.2742.3291.850
(1.848)(1.819)(1.803)(1.770)(1.818)(1.786)
Observations1,7571,7571,7571,7571,7571,757
Correctly classified (%)88.3388.5688.2288.1088.3388.39
Pseudo R20.47470.47600.47240.47120.47420.4747

Note(s): Estimated coefficients (std. errors) using the logit regression of equation (1). The dependent variable is CLEAN in all the regressions. The first row indicates the centrality measure used. All the estimations include year and industry-fixed effects and clustering by firm. See Appendix I for the definition of the variables. ***, **, * for 1%, 5% and 10% significance level

or Create an Account

Close subscription notice
Close access options